The Cardiovascular Healthy Study (CHS) is an NIH-sponsored population-based longitudinal study to identify risk factors for the onset and course of coronary heart disease and stroke in older adults. Holter monitoring was performed on 1,429 participants at baseline and repeated 5 years later in 864. Recordings were also made, at that time, in 385 additional minority participants who entered the study after baseline. In phase 1 of this project, Holter tapes were analyzed, intervals between labeled beats stored, and heart rate variability (HRV), an index of cardiac autonomic function, calculated. In a multivariate model, after 9 years of follow up, only increased age had a stronger association with all-cause mortality than decreased HRV, which was strongly associated with both cardiovascular and, surprisingly, non-cardiovascular mortality. In phase I, ECGs from the Holters were preserved on CDs to permit identification of additional risk factors from ECG waveform measurements. In the phase II, these additional risk factors for outcomes will be extracted. Data will be obtained by measurement of QT-interval-derived and other components of the ECG waveform, detailed analyses of heart rate, HRV, and ECG waveform-derived parameters during sleep periods and upon waking up, and analysis of abnormal heart rate patterns. Detailed analysis of sleep periods should provide information about autonomic patterns, autonomic arousals, respiration, sleep-disordered breathing, time asleep, and naps. It is expected that these analyses will provide descriptive and prognostic information beyond that obtained from standard polysomnography. Analysis of sleep time from Holters should provide a more complete picture of actual sleep time, e.g., sleep in the evening before bedtime, than polysomnography. For subjects in both the Sleep Heart Health Study (SHHS) and the CHS, detailed analyses of sleep periods will be performed on ECG signals obtained during their sleep studies and results directly compared to polysomnography. Results will also be compared with sleep-related questionnaires. Analysis of the SHHS ECGs will permit retrospective analysis of sleep-related changes during 4 nights over up to 14 years in approximately 200 subjects, and the association of these changes with aging and outcomes. Finally, new, highly sophisticated HRV techniques will be applied to the dataset to determine if they provide additional information about sleep and outcomes beyond that from traditional HRV.